Use of Risk-Based Spray Drift Buffers for Protection ... - ACS Publications

Chapter 23

Downloaded by NORTH CAROLINA STATE UNIV on November 8, 2012 | http://pubs.acs.org Publication Date (Web): November 6, 2012 | doi: 10.1021/bk-2012-1111.ch023

Use of Risk-Based Spray Drift Buffers for Protection of Nontarget Areas Scott H. Jackson,*,1 Mark Ledson,2 and Michael Leggett3 1BASF

Crop Protection, 26 Davis Drive, Research Triangle Park, NC 27709 Crop Protection, 410 S. Swing Road, Greensboro, NC 27409 3CropLife America, 1156 15th Street NW, Washington, DC 20005 *E-mail: [email protected].

2Syngeta

Regulatory authorities have begun to shift spray drift management label language from protecting threatened and endangered species and habitat to protecting non-target areas including grasslands, forested areas, shelter belts, woodlots, hedgerows, riparian areas, and shrub lands”. We outline a process for using spray drift models to determine buffer distances using non-target species study endpoints (NOER’s, ER25’s or ER50’s). The conservatively protective nature of this approach is explored, as well as its current use in regulatory practice. Use of the presented method provides adequate protections for current or future habitat. Use of risk-based buffers should be included as part of an accepted toolbox of fixes for non-target species exposure concerns. The U.S. Environmental Protection Agency’s risk assessment and registration processes include spray drift considerations, and approved labeling may include drift reduction considerations. These considerations are based on estimates of potential exposure from drift and hazard evaluation of the chemical being applied. The potential exposure from drift is estimated using models. Stakeholders are currently sponsoring research to improve model estimates so that they more accurately reflect potential for drift with consideration of available technology.

© 2012 American Chemical Society In Pesticide Regulation and the Endangered Species Act; Racke, K., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2012.


Introduction The Office of Pesticide Programs within the U.S. Environmental Protection Agency (EPA) is responsible for administration of the Federal, Insecticide, Fungicide and Rodenticide Act (FIFRA), and has the authority to register plant protection products under the Act. Plant protection products may be registered for specific uses if those uses are deemed, by the Agency, to not represent a risk to the continued well-being of exposed individuals and if they are further deemed to have no unreasonable adverse effects on the environment. To make this latter determination the EPA conducts an ecological risk assessment. Under the provisions of the Endangered Species Act (ESA) all Federal departments and agencies must seek to conserve threatened and endangered species and must insure any “action” does not jeopardize a species or destroy or adversely modify its habitat. Because the “action” of registering or reregistering an active ingredient must be compliant with ESA, the ecological risk assessment conducted by EPA must consider the impacts on endangered species. There are a number of ways that endangered species protections might be applied in plant protection product registration actions. One proposed method is that a more stringent protection standard for ESA would be applied only to the locations identified as habitat for a particular endangered species that could be affected by a proposed product use. In this manner the impact on production agriculture would be minimized by narrowing the land area subject to risk mitigation practices. There are challenges in identifying the range of area that requires protection for each species, potential movement of species and the fact that new species may be added to protection lists requiring revision of restrictions. Additionally, there is disagreement between EPA and the ‘Services’ (U.S. Fish and Wildlife Service and National Marine Fisheries Service) which administer ESA regarding what needs to be considered in an acceptable risk assessment to sufficiently estimate the potential impact on species. Despite obstacles, great strides have been made in developing a system that considers the complexity of questions to be considered. An alternative method has also been applied. EPA has started revising label language to include spray drift protections away from solely occupied threatened and endangered species habitat to protecting all non-target areas identified as grasslands, forested areas, shelter belts, woodlots, hedgerows, riparian areas, and shrub lands. Examples of older label language for spray drift management can be seen in Figure 1, while an example of newer language can be seen in Figure 2. The change to protecting all non-target areas would protect both present and future habitat and remove necessity to consider ‘exclusions’ that may not be static and would provide protections without confirmation of species presence. While agreement on a near zero exposure estimate may be easier to attain than agreement on data requirements for risk assessment, such an approach may cause undue impacts on the grower. To achieve this scenario registrants could apply mitigation measures from an approved toolbox, (e.g., no- spray buffer zones or vegetative buffer strips), to achieve a theoretical ‘de minimus’ exposure scenario. The size of buffers would be based on exposure estimates relative to toxicity endpoints that enable EPA to 326 In Pesticide Regulation and the Endangered Species Act; Racke, K., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2012.


reach a conclusion of “not likely to affect”. Because the risk assessment considers all non-target areas, which include all threatened and endangered species habitat, a “not likely to adversely affect” (NLAA) determination should absolve EPA of the obligation to consult with the Services and could streamline the process. However, achieving a theoretical ‘de minimus’ exposure scenario may have an unacceptably high cost in terms of the amount of land taken out of production to buffer all non-target areas. To achieve an NLAA determination the EPA may choose to use a NOEC (no observed effect concentration) rather than ER25/ER50/ LC50 endpoints. This would likely be substantially more conservative than current assessments and require larger buffers. The work presented here describes a method whereby buffers to protected areas can be calculated using spray deposition models and appropriate non-target effects data. Since the method described in this process is based on exposure calculations and properly selected and applied effects data endpoints, the method is referred to as “risk–based” buffer calculation in contrast to statutory or “expert opinion” buffers which are arbitrarily set without the benefit of scientific method. The paper further explores where improvements may be made to existing tools, using best available data, in order to improve the viability of the methods currently in use or envisioned. Any method applied to this setting must ensure that endangered species protections are adequate while minimizing the impact on production agriculture and the ability to continue to produce food, fiber and fuel for a growing global population.

Methods In order to calculate risk-based buffers, there are two elements required for estimation. The two elements are exposure, which comes from a drift model, and an estimated effects level of concern, which comes from the appropriate nontarget species study. In an example typical for the risk assessment of herbicide application, we have examined effects on non-target plants resulting from ground sprayer exposures. Figure 3 is an illustration of the process required to determine an appropriate risk-based buffer distance. Nontarget Plant Studies For determination of terrestrial spray buffers, data from one of two guideline studies are used which are part of all regulatory data packages. The seedling emergence study OPPTS 850.4225 (1) and the vegetative vigor study OPPTS 850.4250 (2) are part of all regulatory data packages. The decision on which of these studies is used is driven by the product use pattern and the study that demonstrates the greatest sensitivity to the plant protection product being tested. These studies include ten different plant species involving both monocots and dicots. The species included in the studies can vary but typically would be corn, ryegrass, onion, wheat, lettuce, soybean, tomato, cabbage, carrot, and canola. 327 In Pesticide Regulation and the Endangered Species Act; Racke, K., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2012.


Figure 4 is a picture of plants ready for treatment in a vegetative vigor study. For the vegetative vigor study, once plants reach the proper size for treatment, they are sprayed in a spray chamber. Seeds in the seedling emergence study have the product doses placed (or administered) into the soil with water. For the vegetative vigor study, plants are evaluated at 7, 14, and 21 days for plant height, survival and dry weight. Evaluations are similar for the seedling emergence study. Figure 5 is an example table from a vegetative vigor study indicating the non-target plant endpoints. The use of these studies has been criticized for not including weed species. However, the species tested are a cross-section of plant types and testing is done on pre-emergent or very young plants that are most sensitive to plant protection products. It is normal in this testing to see high levels of sensitivity in test species with traits similar to target weed species. In the example table in Figure 5, the no observed effect rate (NOER) for the lowest endpoint used was based on plant height. In the risk assessment process, the effects observed in the laboratory are extrapolated to populations of all non-target plants. When calculating buffer distances, the response level appropriate to achieve a desired protection goal must be known. It is unlikely that an exposure level producing an effect in the laboratory would produce the same level of effect on a heterogeneous population in a field spray scenario. Therefore, the effects tests used represent a very conservative approximation of the effects that might be observed under field conditions, regardless of the level of protection deemed appropriate. The protection goal could be the same for threatened and endangered species as for sensitive non-target crops. An ER25 (25% effect rate) value has often been deemed adequate to provide a margin of safety ensuring that no unreasonable adverse effect would be observed in wild populations exposed at a comparable level. Alternatively, a much more conservative NOER value may be, and often has been, used for assessment.

Spray Drift Exposure Modeling

Spray drift exposure estimates are derived from models of empirical data which are obtained in field studies designed to determine deposition of drift under circumstances prevailing at the time of the study. For FIFRA regulation, two models are currently used predict deposition, AgDRIFT® (3), based on data generated by the Spray Drift Task Force, and AGDISP (4). In Canada, the Pest Management Regulatory Agency (PMRA) uses the Agricultural Buffer Zone Workbook (5) for estimating drift deposition based on data generated by Agriculture and Agri-Food Canada (AAFC) (6). The parameters that can be varied to adjust a buffer distance using available ground models depend on the variables monitored and controlled for in the underlying data. Typically, droplet spectra (VMD50), release height, and wind speed can be changed. Aerial models have many other factors than can be adjusted to impact predicted spray drift. 328 In Pesticide Regulation and the Endangered Species Act; Racke, K., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2012.


329

Figure 1. Example of older label language for the protection of threatened and endangered species. Example is from BASF’s Prowl® H2O label.

In Pesticide Regulation and the Endangered Species Act; Racke, K., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2012.


330 Figure 2. Example of the most recent label language. Protection is wind-directional and has the expanded border areas protected. Example is from BASF’s Kixor® containing product labels.



331 Figure 3. An illustration of the iterative process used for determining risk-based buffer distances. (see color insert)



Figure 4. Plants at about the proper treatment size and growth stage for a vegetative vigor study.

Figure 5. Image of a report table from a vegetative vigor study.

Appropriateness of Model Selection The output of the three models used by North American regulatory agencies, AgDRIFT, AGDISP, and the PMRA tool, are compared in Figure 6, which illustrates how conservative the estimates of the two former models are. Both AgDRIFT and AGDISP deposition curves are for the same spray quality or VMD50 (or nozzle). Both the AgDRIFT and AGDISP deposition curves should be close to the AAFC (Agriculture and Agri-Food Canada) flat fan data, which is consistent with the Spray Drift Task Force Data (7, 8) that the AgDrift model is based on. However, at 400 feet both models greatly over predict deposition. The 332 In Pesticide Regulation and the Endangered Species Act; Racke, K., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2012.


PMRA model provides a good approximation of the AAFC field data assuming use of modern air induction nozzle technology. We can see from this evaluation that both AgDRIFT and AGDISP do not match the field data they are meant to predict, while the PMRA tool provides a much better fit.

Figure 6. Deposition curves from models and field data.

Table I. Comparison of data used for AgDrift ground and the PMRA model 2000 AAFC Datasets

SDTF Datasets

Variable function (3-16 MPH)

No - 1 speed assumed

Sample intervals (edge of field)

3 – 394 ft

26 – 1200 ft

Number of trials

29 usable studies

10 not all usable

Number of years/locations

1 location/2 year

1 location/2 year

Boom heights

2

2

Nozzles types

5 (air induction included)

4 (older types only)

2000, 2004

1992, 1993

Wind

Year Data generated

A comparison of data generated by AAFC and the SDTF are summarized in Table I. While there are still significant differences in the two underlying datasets, the results obtained are very comparable for trials in which the technology used was analogous. Many of the differences in model predictions and the field data they were derived from can be attributed to data analysis and summarization (especially in the case of AgDRIFT). 333 In Pesticide Regulation and the Endangered Species Act; Racke, K., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2012.


Determining Buffer Distances The model estimate is compared to the endpoint chosen as a level of concern. Using the study illustrated in Figure 5, the level of concern could be 0.00026 lb/acre based on soybean plant height effects ((no observed effects rate). This would mean for the purposes of setting a risk based buffer, the spray parameters would need to be adjusted until exposure was less than or equal to 0.00026 lb/acre. If a buffer of 200 feet was deemed to be agronomically viable, the allowable droplet size, maximum wind speed, and boom height could be adjusted until exposure was 0.00026 lb/acre at 200 feet and restrictions would be placed on the label accordingly. Figure 7 is an illustration of attempting to reach a particular goal using AgDRIFT. Using a high boom setting in the model, and the averaged spray quality categories fine to medium/course, the ER25 protection goal was met with a 900 foot buffer. It was not possible to meet the NOER protection goal with the same parameterization. In order to meet the NOER protection goal, spray quality, boom height and wind speed would need to be altered.

Figure 7. Example deposition curve from AgDrift. The blue line is the ER25, while the red line is the NOER. (see color insert) Figure 8 is another comparison of AgDRIFT to field data. This comparison is of model prediction to the SDTF (Spray Drift Task Force) data upon which model development was based. In this comparison, the buffer required for the 0.001 fraction of applied rate would require a distance of 755 feet based on the model, or 150 feet based on the actual deposition data. 334 In Pesticide Regulation and the Endangered Species Act; Racke, K., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2012.


Figure 8. A comparison of Ag DRIFT to Spray Drift Task Force data.

Figure 9. Schematic illustration of the wind tunnel setup used in the Kansas State University study.

335 In Pesticide Regulation and the Endangered Species Act; Racke, K., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2012.

Factors Mitigating Drift and their Consideration in Risk Assessment and Regulation


Nozzle Selection While many factors can affect off-target movement of spray, the factor that has the greatest impact on off-target movement is nozzle selection or droplet size (VMD50). An example of this has been well illustrated by the Kansas State University wind tunnel experiments. A schematic of the wind tunnel setup can be found in Figure 9. The spray was atomized and moved down the chamber. A cross-wind was placed on the spray as it moved down the tunnel, and there was water soluble paper directly across from the source of the cross wind. The percent spray coverage on the water soluble paper is a measure of drift, relatively speaking, that could be expected from each nozzle. The results presented in Figure 10 reflect the percent coverage from the atomized spray that was deflected onto the water soluble paper for the various nozzle types. The flat fan TR8004 (tr8004) nozzle provided 100% coverage on the water soluble paper, while the AI11004 (ttj04) nozzle provided about 10% coverage on the paper. This work indicates how the large differences in nozzles can be used to inhibit or induce spray droplet off-target movement.

Figure 10. Off-target deposition results by nozzle from the Kansas State University study.

As a further illustration of the effect of nozzles in mitigating off target movement, two nozzles were modeled using AGDISP. The results from that modeling can be found in Figure 11. 336 In Pesticide Regulation and the Endangered Species Act; Racke, K., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2012.


Figure 11. Comparison of two nozzles for controlling off-target drift and resulting buffer distances. The results presented in Figure 11 indicate that by proper nozzle selection, spray buffers to protected areas can be minimal or not required at all if nozzle use is considered in risk assessment and stipulated on the label.

Canadian Approach to Buffer Management EPA and the PMRA are analogous regulatory agencies. However, the product use restrictions for a wheat grower in Montana and a wheat grower in Saskatchewan may be very different due to the different way that buffer distances are calculated. The Canadian regulators have developed an online tool (9) that allows users to determine the appropriate level of buffer mitigation given the field conditions and drift-reducing technologies, for example nozzle selection, used at the time of application. This enables Canadian growers to reduce the buffer from the label ‘base case’ distance which is based on worst-case estimates. This approach fosters use of drift-reducing technologies by farmers. It is facilitated by inclusion of air induction and chamber drift-reducing nozzles as mitigation options based on underlying AAFC data that is the most current and best available science. The use of the tool might be extended to including the influence of near-field windbreaks, and other habitat-inducing enhancements in near-field regions, in mitigating drift.

Real Impact of Buffers We have outlined a process for the use of risk-based buffers. One question remaining is whether the implementation of buffers has any real impact on production agriculture. Figures 12 and 13 are GIS images indicating how far into a field a 150 foot buffer and a 250 buffer, respectively, would encroach onto 337 In Pesticide Regulation and the Endangered Species Act; Racke, K., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2012.


agricultural land from the riparian area they are meant to protect. The buffer area could not be sprayed if winds were moving toward the sensitive areas. The use of this buffer area approach would require that land in the buffer area not be sprayed if wind was moving toward the sensitive area, or it would have to be sprayed when the prevailing wind direction was away from the riparian area.

Figure 12. A 150’ buffer to a riparian area. The yellow line is the 150 foot buffer boundary. (see color insert)

One consideration of the use of buffers is the potential disincentive for growers to improve or create habitat in riparian areas resulting in new regions that warrant protection and require additional buffers. The USDA Natural Resource Conservation Service (NRCS) and private conservation groups work diligently to provide incentives to farmers to place marginal lands and near-stream areas under conservation. The most realistic estimates of exposure should be applied in order to minimize the impact of no-spray buffers on production agriculture and ensure that riparian and near-field habitat improvement is fostered. 338 In Pesticide Regulation and the Endangered Species Act; Racke, K., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2012.


Figure 13. A 250’ buffer to a riparian area. The red line is the 250 foot buffer boundary. (see color insert)

Conclusion Protection of threatened and endangered species can be achieved through the use of risk-based buffers. Spray buffers to grasslands, forested areas, shelter belts, woodlots, hedgerows, riparian areas, and shrub lands not only protect current habitat and species, but they may also protect future habitat. However, the riskbased buffer approach requires the use of best available science tools. Currently AgDRIFT is “old science”; a static tool fixed by the technology that described it in the early 1990’s. We need tools that are flexible, and able to accommodate changes in technology (e.g., newer types of spray nozzles). The Canadian approach for calculating buffers is a much more pragmatic process than we currently have with FIFRA (or ESA). However, a software tool framework should be implemented so that as new technology emerges, it can be evaluated, and if deemed appropriate, incorporated into the tool. 339 In Pesticide Regulation and the Endangered Species Act; Racke, K., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2012.

References 1.


2.

3. 4.

5. 6.

7.

8.

9.

USEPA. Ecological Effects Test Guidelines OPPTS 850.4225 Seedling Emergence; Tier II United States Environmental Protection Agency EPA 712–C–96–363, April 1996; http:// www.epa.gov/opptsmnt/pubs/frs/publications/OPPTS_Harmonized/ 850_Ecological_Effects_Test_Guidelines/Drafts/850-4225.pdf. USEPA. Ecological Effects Test Guidelines OPPTS 850.4250 Vegetative Vigor; Tier II United States Environmental Protection Agency EPA 712–C–96–364, April 1996; http://www.epa.gov/ocspp/pubs/frs/publications/OPPTS_Harmonized/ 850_Ecological_Effects_Test_Guidelines/Drafts/850-4250.pdf. AgDRIFT® Spray Drift Taskforce software, Version 2.0; http://www.agdrift.com/AgDRIFt2/DownloadAgDrift2_0.htm. U.S. Department of Agriculture Forest Service. AGDISP U.S. Forest Service spray modeling software; 2006;http://www.continuum-dynamics.com/pragdisp.html. Health Canada. Agricultural Buffer Zone Workbook; 2005; http://www.hcsc.gc.ca/cps-spc/pest/part/consultations/_pro2005-06/index-eng.php. Wolf, T; Cladwell, B. C. Development of a Canadian spray drift model for the determination of buffer zone distances; In Expert Committee on Weeds, Proceedings of the 2001 National Meeting, Quebec City, Sainte Anne de Bellvue, Quebec, 2001; Bernier, D., Campbell, D. R. A., Cloutier, D., Eds.; ECW-CEM: 2001. Johnson, D. Ground Field Study in Texas.. Spray Drift Task Force; Sponsor study number F92-009, MRID No. 43493801 McKenna & Cuneo, 1900 K Street, NW, Washington, DC 20006, Subdivision R- USEPA Pesticide Assessment Guidelines Pesticide Spray Drift Evaluation Guideline 202-1, 1992. Johnson, D. Ground Field Study in Texas Spray Drift Task Force; Sponsor study number F93-016 McKenna & Cuneo, 1900 K Street, NW, Washington, DC 20006, Subdivision R- USEPA Pesticide Assessment Guidelines Pesticide Spray Drift Evaluation Guideline 202-1, 1994. Health Canada. Agency Buffer Zone Calculator; January 2011; http://hc-sc.gc.ca/cps-spc/pest/agri-commerce/drift-derive/calculatorcalculatrice-eng.php.

340 In Pesticide Regulation and the Endangered Species Act; Racke, K., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2012.

Use of Risk-Based Spray Drift Buffers for Protection ... - ACS Publications

Recommend Documents